Darwinism and Marxism

Mike Macnair reviews The structure of evolutionary theory by Stephen Jay Gould (Belknap/Harvard University Press, 2002, pp1,433)

Stephen Jay Gould, who died in May 2002, was widely viewed as a scientific ally of the left. In part this was due to his vigorous public polemics against the christian fundamentalists’ ‘creation science’; against the revived rightwing ‘social Darwinism’ of ‘sociobiology’ in the 70s and more recently ‘evolutionary sociology’; and against pseudo-scientific racism and the cult of IQ tests.

In addition, however, for those Marxists who take the physical sciences seriously, Gould’s most fundamental theoretical idea within evolutionary theory, ‘punctuated equilibrium’, seemed to provide support for our own understandings of the revolutionary transformation of previously stable social orders and the more general dialectical ‘transition of quantity into quality’. Indeed, as Gould recounts in The structure of evolutionary theory, some opponents of ‘punctuated equilibrium’ immediately characterised it as Marxist, and Gould and his co-author, Eldredge, themselves in 1977 linked the dogmatic gradualism of the standard neo-Darwinian model to the ideological gradualism of ‘western’ political ideas (pp984-985, 1017-1019).

In this work Gould offers a systematic account of what he thinks Darwinian evolutionary theory is, how it has developed, what its strengths and weaknesses are, and how it needs to be revised and updated in order to maintain its core insights in the face of our developing knowledge of biological processes and histories. The book is massive: more than 1,400 pages. It is also a lot more dense and technical than his brilliant popular-science books and articles, though Gould’s lucid style still makes it possible for the non-specialist to follow the logic of the argument.

I am not trained as a biologist and hence cannot possibly pass judgment on the extent to which Gould’s account of evolutionary theory is confirmed or undermined by the available experimental evidence, or on the fine detail of the argument. In this review I aim to deal with two issues. The first is that, whether he is right or wrong, Gould’s method of argument in approaching the problem of Darwinism and evolutionary theory is exemplary and can be a partial guide to Marxists approaching the problem of Marxism in the 21st century. The second is that, if Gould is, broadly, right on the nature of the evolutionary theory usable in the 21st century, this, too, has some important substantive lessons for the sort of Marxism which will be useful.

Gould’s structure

The structure of The structure ... is given by Gould’s method for approaching Darwinism in the 21st century. The book opens with a substantial introduction, including a 36-page abstract of the argument. Then part I (pp93-591) deals with the history of Darwinian ideas, the criticisms of them which were in Gould’s view central, and the emergence and ‘hardening’ of the ‘modern synthesis’ of Darwinian natural selection and Mendelian genetics from the 1940s. Part II (pp595-1,343) offers Gould’s version of a “revised and expanded evolutionary theory”.

At first glance it might be thought that there are two books here: one on the history of a science and another on contemporary issues in it. In fact, however, to separate the two parts would cripple the logical structure of Gould’s argument. What part I sets out to do is not to give us a full history of evolutionary biology. Rather, in chapter 2 Gould offers us an explanation of Darwin’s Origin of species as “one long argument”. What were its antecedents and what were its central theoretical commitments - those claims which have to be true if the argument as a whole is to make sense?

Gould starts with the view that Darwin offers an argument against the functionalist creation theory of Paley’s Natural theology, which breaks with the tradition of purely speculative reason about natural history: in contrast, Darwin claims that explanatory hypotheses must be testable against evidence. Darwin draws inferences from the present about the past. Within this framework, Darwin’s argument involves a set of specific commitments which are fundamental to his conclusion that the full range of existing and extinct species can be explained by the process of natural selection. Gould argues that Darwin replaces the creationists’ “divine watchmaker” with the “hidden hand” of the classical economists and especially Adam Smith, the idea of competition between free agents producing functional specialisations in the economy.

But for this to be a satisfactory explanation of biological evolution and the diversity of species, a number of concrete assumptions are needed. The first is that selection operates on individual organisms (successful organisms have more descendants), not on groups of organisms (pp125-27). The second is that selection is strong enough to act as a genuinely dominant creative force(pp137-159). But for this to be the case, variation between individuals must be copious - ie, individuals must be very variable; small in extent - ie, the variations must not generally be so sharp as to create new species immediately; and undirected - ie, the variations must not themselves be driven by functional need (as in Lamarck’s version of evolution) or very sharply constrained or channelled in particular directions by prior structure and history.

Finally, natural selection is driven by the species’ fitness for its environment. But since this is a pretty slow process, in order for selection to produce the range of differences among species that exist today, organisms have to live in the “uniformitarian” world of very gradually changing physical environments. Selection could not be a prime cause sufficient to account for the diversity of species if the history of life on earth was very short, or was punctuated by random mass extinctions (pp159-163). Moreover, substantial population pressure (too many individuals for all of them to do well) is required for competition to play a decisive role; this last element Darwin took from Thomas Malthus’s controversial theory of overpopulation (p122). Darwin is thus profoundly committed to a gradualist account of processes of change.

A striking feature of Gould’s reconstruction of Darwin’s argument is his insistence on focusing on its logical core. At pp147-49 Gould insists that Darwin’s claim for the centrality of natural selection in evolution is a claim for its dominant relative frequency, not one which wholly excludes other mechanisms. This leads him into a sharp polemic against “citation grazing” - the extraction of side and secondary comments out of context to support arguments. The conclusion is worth quoting:

“Proper textual analysis requires that general tenor, not selective statement, be presented. Two basic procedural modes, each with distinctive criteria, set the framework for such textual analysis. The empirical mode makes its judgments of importance by relative frequency and interconnectedness of statements. Meanwhile, and simultaneously, the logical mode employs theoretical consistency as an arbiter for judging the validity and power of the structure of argument. We revere Darwin because he unfailingly manifested the two key traits of brilliance and honesty. He knew where his arguments led, and he followed them relentlessly, however unpleasant the consequences. We do him the greatest possible disservice when we approach his work as a superficial grazer, searching for some particular item of personal sustenance, while ignoring the beauty and power of general tenor and logical entailment” (p149).

Chapters 3-6 explore challenges, alternatives and modifications to Darwinian theory in the later 19th and early 21st century and the rise of the ‘modern synthesis’. The approach is structured by the points of the core logic of Darwin’s argument which Gould has previously identified: chapter 3 discusses ideas of selection or other processes of change taking place above the level of the individual organism (species or populations) or below it (‘germ plasm’, etc); chapters 4 and 5 explore structural, as opposed to functional, causes in the internal logics of biological forms; chapter 6 difficulties caused by debates over the age of the earth and geology. Gould’s detailed exploration of these issues helps us to understand the narrative in chapter 7 of the rise and gradual hardening of the ‘modern synthesis’, with increasing emphasis put on naturally selected adaptation at the expense of other causes of change, on reduction to the individual or genetic level, and on gradualism. Part I thus sets the stage for part II, which is Gould’s exploration of how, in his view, Darwin’s argument must be modified.

As will by now be no surprise, the same themes - the core logical elements of Darwinism - resurface. Chapter 8 presents a long and elaborate argument for the view that species, as well as individuals, can be naturally selected in the Darwinian sense, and a repeat and development of Gould’s earlier arguments against the ‘selfish gene’ reductionism of Dawkins and others. Gould argues that it is possible to distinguish six levels at which evolution may operate: gene, cell, organism, deme (local population), species and clade; and that at each level the relative weights of selection for adaptation, pure chance, internal constraint and limitation from the higher levels are different. Chapter 9 is an extended defence in this context of ‘punctuated equilibrium’, particularly emphasising the testability of the idea against opponents who have argued it is untestable, and its implications for teleological versions of Darwinism.

Chapter 10 focuses on structural and historical constraints or ‘channels’ of possible variations, drawing both directly on physical constraints on possible forms, and for the bulk of the chapter on recent discoveries in genetics made possible by DNA sequencing. These have revealed that some instances of ‘convergent evolution’ between very different species, which have been thought to be purely driven by adaptation, turn out to involve common genetic switches which build the structure of the organism in the course of its development. Chapter 11 elaborates further on the theme of structural constraints, and on ideas developed earlier in Gould’s work: ‘spandrels’ or non-functional consequences of functional architecture; and ‘exaptations’ or characteristics which are originally either adaptations for other purposes or spandrels, but which become adaptive under changed circumstances.

The final chapter, chapter 12, returns to the question of time and catastrophism via mass extinctions, like that of the dinosaurs, and the late 20th century discovery that this was the chance result of an asteroid impact rather than resulting from any maladaptation of the dinosaurs as such. As with multiple levels of evolution, Gould now argues that it is also necessary to deal in multiple “tiers of time”: causal processes operative at the micro-level of centuries may not have the same weight at the macro-level of millennia.

This is a bald and grossly oversimplified summary of the gist of the argument - remember that Gould himself required 36 pages for the abstract! It also omits not only the technical detail, but also the fascinating examples, flights of rhetoric and analogies which make the book, long and tough as it is, an enjoyable read. But these ugly bare bones possibly allow us to draw some lessons from Gould’s approach to evolutionary Darwinism, for how we need to approach Marxism.

Lessons for Marxism

Why should Marxism take lessons in method from evolutionary biology? The most basic reason is that what Marx and Engels thought they had founded was a scientific socialism. They claimed to have departed from the realm of speculative reason into that of the testable - like Darwin’s Origin of species, which they welcomed. (A variety of New Leftists and academic ‘Marxologists’ in the 1960s and 1970s claimed to be able to detect differences on this issue between Marx and Engels. These claims have been sharply and unanswerably refuted from the evidence of Marx and Engels’s working methods by Sebastiano Timpanaro in his On materialism and by Hal Draper in Karl Marx’s theory of revolution Vol 1.) More generally, if Marxism is not scientifically acceptable - that is, if, for example, ‘positive political science’, ‘evolutionary sociology’, or marginal-utility economic theory were shown to have more scientific explanatory and predictive power than the core theses of Marxism - then promoting Marxism in politics would be just as immoral as selling cars without brakes, on the basis that, following Aristotelian mechanics, switching off the engine would be enough to stop the car.

Marxism as a scientific theory of political action faces markedly similar problems to Darwinism as a scientific theory of evolution. Much of the evidence is in the past and thus not susceptible to experimental manipulation. The laws asserted are laws of tendency, not absolute laws; what is claimed for their operation is dominant relative frequency, not absolute mechanical effectiveness. The original texts proposing the theory were written more than a century ago and were interventions in debates of that period. There has been an extensive subsequent literature which has challenged the theoretical foundations as well as the detail. And there has been a considerable accumulation since then of both relevant historical evidence (produced by historians, archaeologists and anthropologists) and experimental evidence (produced by the workers’ movement and parties and organisations attempting to use Marxism in political decision-making).

Equally, since around the aftermath of the Russian Revolution, Marxism has been tragically affected by an extreme form of “citation grazing”, in which the works of the founders are treated as a sort of Holy Writ to be ‘mined’ for authority on this or that point of detail. What ‘counts as’ Marxism is even more debateable than what ‘counts as’ Darwinism.

In this context, Gould’s method is one from which Marxists can definitely learn. Doing Marxism in the 21st century requires us to find an alternative to “citation grazing”. Gould’s method of approach to Darwinism - analysing the core logic of the approach, how it has been challenged or modified since, and how both the original arguments and these challenges or modifications hold up in the face of the whole body of the evidence - is clearly the right method of approach to Marxism. Now this may sound at first like a repeat of the project of the ‘analytical Marxists’, and it is indeed what they pretended to do (GA Cohen Karl Marx’s theory of history: a defence [Oxford 1978] is the ‘founding text’ of this academic tendency; J Roemer [ed] Analytical Marxism [Cambridge 1986] contains a variety of essays within the school).

In reality, however, it was not what they actually did. Rather, the analytical Marxist school started with a silent and wholly unexamined commitment to the (manifestly disproved) theses of the 7th Congress of the Comintern on people’s fronts and ‘national roads to socialism’, elevating these into a methodological commitment to analysing social formations exclusively at national level. On this basis they went on to accept uncritically a whole range of standard academic criticisms of Marxism and to attempt to ‘reconstruct’ Marx in a way which was not vulnerable to these criticisms. The ultimate collapse of this project into ordinary liberalism or ‘ethical socialism’ was inevitable, given its anti-scientific starting point in ignoring evidence inconsistent with their methodological nationalism. (The collapse is discussed by Marcus Roberts [Analytical Marxism: a critique Verso 1996] who offers a ‘post-Marxist’ critique while maintaining the assumptions of the ‘analytical’ school; and by Daniel Bensaid [Marx for our Times Verso 2002], which reads Marx as a ‘postmodernist’ but does at least take scientific issues more seriously.)

To put the point another way: what is needed is an approach to Marxism which:

If we treat evidence and argument produced by non-Marxist writers as valueless we act uncritically and hence unscientifically, just as much as if we accept it uncritically. It will return to haunt us, just as the proto-Alliance for Workers’ Liberty’s refusal to engage with ‘identity politics’ in the 70s has returned to haunt it as an uncritical acceptance of a particular variant of jewish ‘identity politics’ in the later 80s and 90s. Equally, if we treat any moment, whether it be August 1914 or September 1933, as one of ‘original sin’, after which the attempts of workers’ parties to engage in political action no longer count as evidence in the analysis of Marxism because they are ‘social-democratic’, ‘Stalinist’, ‘Maoist’, ‘Trotskyite’ or ‘Pabloite’, we fail to take into account all the evidence, and in practice will end up repeating old mistakes. Our problem is thus not simply to recover classical Marxism: this approach seriously risks leading to more “citation grazing”.

The key here is programme and strategic line of march. Only if we formulate a programme as a strategic line of march can we subsequently correct and modify it. Where programme is collapsed into general theory and the ‘need for the party’, as it is by the Socialist Workers Party, there is no experiment and no self-criticism, and we get not a scientific Marxism, but a mere abstract dogma maintained by a sect which ‘forgets nothing and learns nothing’. Where it is collapsed into “the need to resist”, as it is by the United Secretariat of the Fourth International (for example, Bensaid’s conversion of Marxism into a pure critique of the world-as-it-is, without testable implications, in Marx for our times), we get merely empirical thrashing around.

It should be apparent that a project of this sort for a Marxism for the 21st century can only be a collective project. Gould’s massive attempt to restate evolutionary theory is nonetheless part of an ongoing debate in the field and draws heavily on the contributions of other theorists. A reconstructed Marxism would have to draw on an even wider range of arguments and evidence. Our lack of a proper party and possession instead of a series of sects means that none of us have the collective resources to engage effectively in this project. The academic semi-Marxist left is no substitute, because it is if anything more sectarian than the far-left sects.

Two examples

I want to end with some points in which Gould seems to me to have substantive things to say relevant to the project of a reconstructed Marxism. These are the ideas of a hierarchy of evolutionary levels, and of the deep-level persistence of constraints of history and form.

The hierarchy of levels is an idea which we can use effectively to combat a very common false choice. The analytical Marxists insisted explicitly on methodological individualism: ie, that social entities should be reducible to aggregates of choices made by individuals (though they also showed an underlying silent methodological nationalism, as I pointed out earlier). The result is a collapse into political individualism and liberalism. Their critics have very commonly resorted either to Hegelian versions of the dialectic, in which the totality is ultimately determinant, or to the ‘structural over-determinations’, which Louis Althusser took from Claude Levi-Strauss. The result is that agency - the ability of classes and individuals to change things - disappears from the explanatory system or comes in (as in Bensaid’s Marx for our times) only as an unexplained external resistance to the system.

We can resist this false choice between methodological individualism and methodological collectivism, if we understand that there is a hierarchy of levels in social ordering and change, that within processes of change on a world scale there are also processes of change which are specific to nations and not fully determined by world developments, and that within these there are choices made by classes and by individuals, and that there is not full mechanical determination of any of these levels by any other. The point could be put the other way round: classes and nations are not reducible to aggregates of individuals, but have ‘emergent’ properties; the world economy/world order is not reducible to an aggregate of nations, but has (different) ‘emergent’ properties. We do not have to choose between ‘methodological individualism’ and an equally sterile system of over-determinations.

The idea of deep historical and formal constraints is again helpful in dealing with two symmetrical errors. A number of social institutions - nation, family and the oppression of women, state, law, etc - long antedate the emergence of capitalism. There have been two general ‘Marxist’ responses to this circumstance. The first, traditionally popular on the far left, is to explain these institutions by their functionality for capitalism as a social system. This is a variant on structural or dialectical over-determination theory; it is also strikingly like the neo-Darwinian functional explanation of all biological features, against which Gould argues. The second, Eurocommunist (etc) version is to treat them as having bases and dynamics completely independent of the class system - hence the ‘multiple oppressions’ approach.

The answer must surely be that while it is hypothetically possible that there could be a capitalism without, for example, the family and the oppression of women, this is no more than what Marx scornfully called a “Robinsonade”, after Robinson Crusoe - a capitalist utopia. Real, existing capitalism grew out of real, existing (European and Japanese) feudalism, which grew out of real, existing pre-feudal social orders, which incorporated the family and the oppression of women. Indeed, it is possible that the social institution of class and its dynamics are meaningless without the inheritance of class position, which inherently involves the family.

Certainly the real, existing capitalist class system necessarily involves elements which in their origin are pre-capitalist, and are not necessarily functional for capitalism - just as, on Gould’s account, real evolution involves not just functional adaptation, but functional adaptation which builds on historical constraints.

Mike Macnair